Mass spectrometry-based quantification of transporter proteins and metabolizing enzymes: an update on advantages and challenges

Size: px

Start display at page:

Download "Mass spectrometry-based quantification of transporter proteins and metabolizing enzymes: an update on advantages and challenges"

Magdalen Ford
6 years ago
Views:

proteins and metabolizing enzymes: an update on

40 30 20 10 ABCB1 (P-gp) 400 200 0 D J1 J2 I C L

1 Protein amount [fmol/mg tissue] Mass spectrometry-based quantification of transporter proteins and metabolizing enzymes: an update on advantages and challenges 600 CYP3A ABCB1 (P-gp) D J1 J2 I C L 0 * D J1 J2 I C L Stefan Oswald, PhD Department of Clinical Pharmacology Center of Drug Absorption and Transport (C_DAT) University Medicine of Greifswald, Germany

2 Clinically relevant transporters The International Transporter Consortium, Nature Reviews Drug Discovery 2010/ Clin Pharmacol Ther 2013

/ transporters To estimate / predict the oral absorption of drugs To estimate / predict the potential

3 Why do we need intestinal expression data? To know which transporters are expressed in a certain tissue To know the absolute abundance of enzymes / transporters To estimate / predict the oral absorption of drugs To estimate / predict the potential contribution of each protein to drugdrug / drug-food interactions source: UCSF-FDA TransPortal To characterize cellular transporter models or animal models GastroPlus

Elimination hepatic Elimination Many drugs are characterized by low and highly

4 First pass-route of drugs Absorption Distribution Metabolism Elimination portal vein liver BIOAVAILABILITY Gut wall Problems of oral absorption fecal Excretion intestinal Elimination hepatic Elimination Many drugs are characterized by low and highly variable oral bioavailability Drug-drug interactions Food-drug interactions renal Excretion

5 phase I / II metabolism Intestinal metabolism and transport small intestinal expression enterocyte P-gp / BCRP ASBT P-gp PEPT1 MRP1 MRP2 Merging of data from different donors Small number of genes / intestinal samples BCRP PEPT1 MRP3 MRP2/3 OATP2B1 OCT1/3 Tissue from segments assessable by biopsies OSTα/ OCT1 OSTβ (duodenum, terminal ileum, colon) OCT3 OATP2B1 Samples from patients ASBT Mostly mrna expression intestinal lumen dataportal vein distal proximal Müller et al. 2017, Update on intestinal transporters (Biol. Chem; 398(2):175-92)

6 Limitations of mrna data mrna expression is not necessarily correlated to protein expression! ABCB1 ABCC2 mrna-arrays Oswald et al. 2013, AAPS Journal: 15(4):

7 Paine et al. 2006, DMD All available protein data were generated via immunoblotting Mouly et al. 2003, Pharm Res BCRP Berggreen et al. 2007, Mol Pharm Tucker et al. 2012, Biochem Pharmacol

Limitations of immunoblotting Impact of rifampin (600

P-gp / MDR1 before rifampin after rifampin protein mrna

5-fold 4.3-fold Oswald et al. 2006, CPT 1.4- / 3.

8 Limitations of immunoblotting Impact of rifampin (600 mg, 6-8 d) on duodenal mrna and protein expression of P-gp / MDR1 before rifampin after rifampin protein mrna reference 8.3-fold 3-fold Giessmann et al. 2004, CPT 1.5-fold 4.3-fold Oswald et al. 2006, CPT 1.4- / 3.5*-fold - Greiner et al. 1999, JCI 1.1- / 4.2*-fold 2.4-fold Westphal et al. 2000, CPT *Western blotting Question: Are the observed differences real or due tue methological issues?

9 Available protein data on intestinal metabolizing enzymes Riches et al Paine et al Shimada et al. 1994

10 analytical signal (optical density, AU) measured protein content (ng) Limitations of immunoblotting 1. Antibody required (availability?, expensive) 2. Functionality uncertain (specificity?) 3. Only few protein simultaneously detectable 4. Poor reproducibility 5. Semiquantitative protein determination 6. Data about analytical quality often not available (e.g. limit of quantification, accuracy, precision) Oswald et al. 2013, AAPS Journal: 15(4): ,5x10 5 2,0x10 5 1,5x10 5 1,0x10 5 5,0x10 4 0,0-5,0x10 4 HSA in PSB ,5x10 5 OATP1B1 3,0x10 5 OATP1B3 OATP2B1 prepared protein amount (ng) prepared protein amount (ng) protein OATP1B1 OATP1B3 OATP2B1 range (µg) , outliner (%) 16,6 20,8 13,3 CV (%) 9,6-94,6 38,5-70,7 33,6-87,8 rel. error (%) -33, ,1-108

and wide analytical range High throughput and automatable

11 Promises of LC-MS/MS based protein quantification High specificity High accuracy and precision High sensitivity and wide analytical range High throughput and automatable technique Analytical quality parameters assessable ( validation)

12 Mass spectrometry based protein quantification: applications o Dass C et al. 1989: beta endorphine o Barr JR et al. 1996: apolipoprotein A-I o Gerber SA et al. 2003: yeast proteins o Barnidge DR et al. 2003: rhodopsine o Barnidge DR et al. 2004: PSA o Anderson L et al. 2006: 53 plasma proteins o Since 2008: transporters, enzymes (established groups: Terasaki lab (Japan), Smith lab (USA), Unadkat / Prasad lab (USA), Oswald lab (Germany), Artursson lab (Sweden))

13 Mass spectrometry based protein quantification: applications

14 Mass spectrometry based protein quantification Protein is enzymatically digested (trypsin) proteospecific peptides Peptides are separated via a HPLC and detected with mass spectrometry Protein identification / quantification by MRM experiment y3 y2 y1 Multiple Reaction Monitoring (MRM) R 1 O R 2 O R 3 O R 4 O H2N-CH C NH CH C NH CH C NH CH-COH b1 b2 b3 R1 H2N-CH R2 O O + C NH CH C + NH3 R3 CH O R4 O C NH CH-COH Ion selection Fragmentation Fragment selection b 2 y 2

Mass spectrometry based protein quantification Label-free absolute quantification ABCB1 (P-gp) CYP3A4 Targeted proteomics QconCAT proteomics Global proteomics (shotgun approach) Absolute

15 Mass spectrometry based protein quantification Label-free absolute quantification ABCB1 (P-gp) CYP3A4 Targeted proteomics QconCAT proteomics Global proteomics (shotgun approach) Absolute quantification of pre-defined proteins using proteospecific peptides and stable-labelled internal standard peptides (~1-50 proteins) Absolute / relative quantification of proteins without internal standards using protein databases (thousands of proteins) General assumption: amount of peptides equal the amount of proteins (complete digestion)

16 Targeted or global proteomics? targeted analysis vs. non-targeted screening What is in the box? targeted analysis: Grape? Apple? Cherry? Banana?

17 Targeted or global proteomics? targeted analysis vs. non-targeted screening What is in the box? targeted analysis: Grape? Apple? Cherry? Banana? no no no yes

18 Targeted or global proteomics? targeted analysis vs. non-targeted screening What is in the box? Non-targeted screening: M I N I O N S H B E L K R T D T A B D E K O G U N C P J G F I M A W V C V F W L N A D H Z A T X A

19 Targeted or global proteomics? targeted analysis vs. non-targeted screening What is in the box? Non-targeted screening: M I N I O N S H B E L K R T D T A B D E K O G U N C P J G F I M A W V C V F W L N A D H Z A T X A

20 Targeted or global proteomics? targeted analysis vs. non-targeted screening What is in the box? Non-targeted screening: M I N I O N S H B E L K R T D T A B D E K O G U N C P J G F I M A W V C V F W L N A D H Z A T X A

21 Targeted or global proteomics? targeted Only expected peptides detected No modifications detected Absolute quantification possible Validation possible (accuracy, precision, stability) High sensitivity & accuracy Non-targeted (global) Vast amount of present peptides detected modifications detectable Relative quantification Validation not possible Low abundant proteins may not be detected Fair Throughput (run time ~1 h) Low Throughput (run time ~3 h)

possible (accuracy, precision, stability) High sensitivity & accuracy Non-targeted (global) Vast amount of

22 Targeted or global proteomics? targeted Only expected peptides detected No modifications detected Absolute quantification possible Validation possible (accuracy, precision, stability) High sensitivity & accuracy Non-targeted (global) Vast amount of present peptides detected modifications detectable Relative quantification Validation not possible Low abundant proteins may not be detected Fair Throughput (run time ~1 h) Low Throughput (run time ~3 h)

measured concentration (nmol/l) LC-MS/MS-based protein quantification 25 20 15 10 5 range: 0.1-25 nmol/l precision: 3.7-12.9% rel. error: -9.3 8.

23 measured concentration (nmol/l) LC-MS/MS-based protein quantification range: nmol/l precision: % rel. error: % spiked concentration (nmol/l) Available Assays Transporter Enzymes ABCB1 CYP1A2 ABCC1 CYP2B6 ABCC2 CYP2C8 ABCC3 CYP2C9 ABCC4 CYP2C19 ABCG2 CYP2D6 ASBT CYP2E1 MATE1 CYP2J2 Na/K-ATPase CYP3A4 NTCP CYP3A7 OAT1-3 CYP4A11 OATP1A2 CYP4F2 OATP1B1 / 1B3 UGT1A1 OATP2B1 UGT1A3 OCT1-3 UGT2B7 OCTN2 UGT2B15 PEPT1-2 UGT2B17 modified according to Oswald et al. 2013, AAPS Journal

24 Critical aspects of mass spectrometry based proteomics Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

25 Critical aspects of mass spectrometry based proteomics - Method Development - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

26 Critical aspects of mass spectrometry based proteomics - Method Development - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

27 Protein sequence databases Critical aspects of mass spectrometry based proteomics - Method Development - in silico digestion

CYP3A4 (503 AA) Critical aspects of mass spectrometry based proteomics - Method Development - MALIPDLAMETWLLLAVSLVLLYLYG THSHGLFKKLGIPGPTPLPFLGNILS YHKGFCMFDMECHKKYGKVWGFY DGQQPVLAITDPDMIKTVLVKECYSV

28 CYP3A4 (503 AA) Critical aspects of mass spectrometry based proteomics - Method Development - MALIPDLAMETWLLLAVSLVLLYLYG THSHGLFKKLGIPGPTPLPFLGNILS YHKGFCMFDMECHKKYGKVWGFY DGQQPVLAITDPDMIKTVLVKECYSV FTNRRPFGPVGFMKSAISIAEDEEW KRLRSLLSPTFTSGKLKEMVPIIAQY GDVLVRNLRREAETGKPVTLKDVFG AYSMDVITSTSFGVNIDSLNNPQDPF VENTKKLLRFDFLDPFFLSITVFPFLI PILEVLNICVFPREVTNFLRKSVKRM KESRLEDTQKHRVDFLQLMIDSQNS KETESHKALSDLELVAQSIIFIFAGYE TTSSVLSFIMYELATHPDVQQKLQE EIDAVLPNKAPPTYDTVLQMEYLDM VVNETLRLFPIAMRLERVCKKDVEIN GMFIPKGVVVMIPSYALHRDPKYWT EPEKFLPERFSKKNKDNIDPYIYTPF GSGPRNCIGMRFALMNMKLALIRVL QNFSFKPCKETQIPLKLSLGGLLQPE KPVVLKVESRDGTVSGA Selection criteria for peptides Mass range of MS (~ ), ~7-25 AA Not localized in transmembrane region (transporter) No genetic polymorphisms (> 1%) No posttranslational modifications No repeated sequence of arginine and lysine No cysteine, methionine or tryptophan Protein (species) specific (BLAST search)

29 CYP3A4 (503 AA) MALIPDLAMETWLLLAVSLVLLYLYG THSHGLFKKLGIPGPTPLPFLGNILS YHKGFCMFDMECHKKYGKVWGFY DGQQPVLAITDPDMIKTVLVKECYSV FTNRRPFGPVGFMKSAISIAEDEEW KRLRSLLSPTFTSGKLKEMVPIIAQY GDVLVRNLRREAETGKPVTLKDVFG AYSMDVITSTSFGVNIDSLNNPQDPF VENTKKLLRFDFLDPFFLSITVFPFLI PILEVLNICVFPREVTNFLRKSVKRM KESRLEDTQKHRVDFLQLMIDSQNS KETESHKALSDLELVAQSIIFIFAGYE TTSSVLSFIMYELATHPDVQQKLQE EIDAVLPNKAPPTYDTVLQMEYLDM VVNETLRLFPIAMRLERVCKKDVEIN GMFIPKGVVVMIPSYALHRDPKYWT EPEKFLPERFSKKNKDNIDPYIYTPF GSGPRNCIGMRFALMNMKLALIRVL QNFSFKPCKETQIPLKLSLGGLLQPE KPVVLKVESRDGTVSGA Critical aspects of mass spectrometry based proteomics - Method Development - mass position variants peptide sequence :D->H 185:T->S 189:F->S DVFGAYSMDVITSTSFGVNIDSLNNPQDPFVENTK :L->P ALIPDLAMETWLLLAVSLVLLYLYGTHSHGLFK :P->R 222:S->P FDFLDPFFLSITVFPFLIPILEVLNICVFPR :T->N 363:T->M APPTYDTVLQMEYLDMVVNETLR VWGFYDGQQPVLAITDPDMIK LGIPGPTPLPFLGNILSYHK (Phosphorylated?, 74%) :I->T DNIDPYIYTPFGSGPR (Phosphorylated?, ~80%) EMVPIIAQYGDVLVR LSLGGLLQPEKPVVLK VDFLQLMIDSQNSK GVVVMIPSYALHR :I->V SAISIAEDEEWK LQEEIDAVLPNK :G->D GFCMFDMECHK :P->S VLQNFSFKPCK DVEINGMFIPK :V->I EAETGKPVTLK SLLSPTFTSGK RPFGPVGFMK ECYSVFTNR YWTEPEK EVTNFLR FALMNMK LFPIAMR ETQIPLK LEDTQK ETESHK :M->T NCIGMR DGTVSGA

30 OCT1 (554 AA) Critical aspects of mass spectrometry based proteomics - Method Development - mass position modifications variants peptide sequence SHLPLGPCQDGWVYDTPGSSIVTEFNLVCADSWK MPTVDDILEQVGESGWFQKQAFLIL CLLSAAFAPICVGIVFLGFTPDHHCQ SPGVAELSQRCGWSPAEELNYTVP GLGPAGEAFLGQCRRYEVDWNQSA LSCVDPLASLATNRSHLPLGPCQDG WVYDTPGSSIVTEFNLVCADSWKLD LFQSCLNAGFLFGSLGVGYFADRFG RKLCLLGTVLVNAVSGVLMAFSPNY MSMLLFRLLQGLVSKGNWMAGYTLI TEFVGSGSRRTVAIMYQMAFTVGLV ALTGLAYALPHWRWLQLAVSLPTFLF LLYYWCVPESPRWLLSQKRNTEAIK IMDHIAQKNGKLPPADLKMLSLEEDV TEKLSPSFADLFRTPRLRKRTFILMY LWFTDSVLYQGLILHMGATSGNLYL DFLYSALVEIPGAFIALITIDRVGRIYP MAMSNLLAGAACLVMIFISPDLHWL NIIIMCVGRMGITIAIQMICLVNAELYP TFVRNLGVMVCSSLCDIGGIITPFIVF RLREVWQALPLILFAVLGLLAAGVTL LLPETKGVALPETMKDAENLGRKAK PKENTIYLKVQTSEPSGT :S->L LCLLGTVLVNAVSGVLMAFSPNYMSMLLFR TVAIMYQMAFTVGLVALTGLAYALPHWR EVWQALPLILFAVLGLLAAGVTLLLPETK :P->L 287:R->G 85:L->F 88:C->R WLQLAVSLPTFLFLLYYWCVPESPR CGWSPAEELNYTVPGLGPAGEAFLGQCR :L->F LDLFQSCLNAGFLFGSLGVGYFADR :M->I 461:V->I MGITIAIQMICLVNAELYPTFVR :G->R NLGVMVCSSLCDIGGIITPFIVFR YEVDWNQSALSCVDPLASLATNR :S->F MPTVDDILEQVGESGWFQK :G->V GNWMAGYTLITEFVGSGSR MLSLEEDVTEK PHOS: 333 LSPSFADLFR IMDHIAQK GVALPETMK VQTSEPSGT PHOS: 541 ENTIYLK LLQGLVSK WLLSQK DAENLGR LPPADLK NTEAIK

31 Critical aspects of mass spectrometry based proteomics - Method Development - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

32 Critical aspects of mass spectrometry based proteomics - Method Development - HPLC Column Ionisation source Mass spectrometer(ms) Data analysis

Critical aspects of mass spectrometry based proteomics - Method Development - Mass transitions are not measured simultaneously but consecutively not unlimited Conventional Time-scheduled MRM run

33 Critical aspects of mass spectrometry based proteomics - Method Development - Mass transitions are not measured simultaneously but consecutively not unlimited Conventional Time-scheduled MRM run Gallien et al., J Mass. Spectrom proteins peptides / protein transitions / peptide sum of transitions* cycle time (s) *including 20 internal standard 3 peptides; assumed 3 dwell time: ms 10.8

34 Critical aspects of mass spectrometry based proteomics - Method Development - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

35 Method validation Important parameters: specificity, linearity, range, LLOQ, accuracy and precision (within- and between-day), stability (short-term, post-preparative, long-term, freeze-and-thaw), matrix effects

36 Intensity, cps Method validation: specificity XIC of +MRM (78 pairs): / amu Expected RT: 51.7 ID: UGT1A3* from Sample 26 (Darm 009b) of Darmproben & Verdu... Max. 1.5e5 cps. 3.6e6 3.4e6 3.2e6 C calibration sample in HSA CYP2E UGT1A e6 2.8e6 2.6e6 In te n s ity, c p s 2.4e6 2.2e6 2.0e6 1.8e6 1.6e6 1.4e6 1.2e6 1.0e6 CYP1A UGT2B CYP2D CYP3A UGT1A CYP3A CYP2B e5 6.0e5 CYP2C CYP2C e5 2.0e5 CYP2C UGT2B Time, min Time, min Gröer et al. 2014, J Pharm Biomed Anal

37 Method validation: specificity Protein Peptide mass Q1 Q3.1 Q3.2 Q3.3 CYP1A2 YLPNPNALQR YLPNPNALQR* CYP3A4 EVTNFLR EVTNFLR* CYP3A5 DTINFLSK DTINFLSK* CYP2B6 TAEIPFSLGK TAEIPFSLGK* CYP2C19 GHFPLAER GHFPLAER* CYP2C8 VQEEIDHVIGR VQEEIDHVIGR* CYP2C9 GIFPLAER GIFPLAER* CYP2D6 DIEVQGFR DIEVQGFR* CYP2E1 FITLVPSNLPHEATR FITLVPSNLPHEATR* UGT1A1 TYPVPFQR TYPVPFQR* UGT1A3 YLSIPTVFFLR YLSIPTVFFLR* UGT2B7 IEIYPTSLTK IEIYPTSLTK* UGT2B15 SVINDPVYK SVINDPVYK* Gröer et al. 2014, J Pharm Biomed Anal

38 Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1_1" Mass(es): "504.6/372.5 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 5.00 ng/ml Calculated Conc: 5.19 ng/ml Acq. Date: 11/21/2013 Acq. Time: 3:55:50 PM 1.25e4 Modified: No Proc. Algorithm: Analyst Classic Bunching Factor: 1 Noise Threshold: cps Area Threshold: cps Num. Smooths: 7 Sep. Width: 0.20 Sep. Height: 0.01 Exp. Peak Ratio: 5.00 Exp. Adj. Ratio: 4.00 Exp. Val. Ratio: 3.00 RT Window: 30.0 sec Expected RT: 9.56 min Use Relative RT: No Int. Type: Base To Base Retention Time: 9.51 min Area: 1.14e+005 counts Height: 1.28e+004 cps Start Time: 9.34 min End Time: 9.77 min 1.20e4 1.15e4 1.10e4 1.05e4 1.00e Time, min Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1*1(IS)" Mass(es): "510.1/377.7 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 1.00 ng/ml Calculated Conc: N/A Acq. Date: 11/21/ e4 Acq. Time: 3:55:50 PM 3.9e4 Modified: No Proc. Algorithm: Analyst Classic 3.8e4 Bunching Factor: 1 Noise Threshold: cps Area Threshold: cps 3.7e4 Num. Smooths: 7 Sep. Width: e4 Sep. Height: 0.01 Exp. Peak Ratio: e4 Exp. Adj. Ratio: 4.00 Exp. Val. Ratio: e4 RT Window: 30.0 sec Expected RT: 9.56 min 3.3e4 Use Relative RT: No Int. Type: Base To Base Retention Time: 9.52 min Area: 3.48e+005 counts Height: 4.06e+004 cps Start Time: 9.34 min End Time: 9.79 min 3.2e4 3.1e4 3.0e4 2.9e4 2.8e4 2.7e4 2.6e4 2.5e4 2.4e4 2.3e4 2.2e4 2.1e4 2.0e4 1.9e4 1.8e4 1.7e4 1.6e4 1.5e4 1.4e4 1.3e4 1.2e4 1.1e4 1.0e Time, min 9.52 Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1_2" Mass(es): "504.6/547.6 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 5.00 ng/ml Calculated Conc: 5.14 ng/ml Acq. Date: 11/21/ Acq. Time: 3:55:50 PM 7400 Modified: No Proc. Algorithm: Analyst Classic 7200 Bunching Factor: 1 Noise Threshold: cps Area Threshold: cps 7000 Num. Smooths: 7 Sep. Width: Sep. Height: 0.01 Exp. Peak Ratio: Exp. Adj. Ratio: 4.00 Exp. Val. Ratio: 3.00 RT Window: 30.0 sec 6400 Expected RT: 9.56 min Use Relative RT: No 6200 Int. Type: Base To Base Retention Time: 9.51 min Area: 7.05e+004 counts Height: 7.72e+003 cps Start Time: 9.33 min End Time: 9.77 min Time, min 9.51 Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1*2(IS)" Mass(es): "510.1/557.7 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 1.00 ng/ml 2.4e4 Calculated Conc: N/A Acq. Date: 11/21/ e4 Acq. Time: 3:55:50 PM 2.3e4 Modified: No Proc. Algorithm: Analyst Classic 2.2e4 Bunching Factor: 1 Noise Threshold: cps 2.2e4 Area Threshold: cps Num. Smooths: 7 2.1e4 Sep. Width: e4 Sep. Height: 0.01 Exp. Peak Ratio: e4 Exp. Adj. Ratio: 4.00 Exp. Val. Ratio: e4 RT Window: 30.0 sec Expected RT: 9.55 min 1.9e4 Use Relative RT: No 1.9e4 Int. Type: Base To Base Retention Time: 9.49 min 1.8e4 Area: 2.20e+005 counts Height: 2.35e+004 cps 1.8e4 Start Time: 9.32 min End Time: 9.79 min 1.7e4 1.7e4 1.6e4 1.6e4 1.5e4 1.5e4 1.4e4 1.4e4 1.3e4 1.3e4 1.2e4 1.2e4 1.1e4 1.1e4 1.0e Time, min 9.49 Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1_3" Mass(es): "504.6/743.8 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 5.00 ng/ml 5100 Calculated Conc: 5.21 ng/ml Acq. Date: 11/21/ Acq. Time: 3:55:50 PM 4900 Modified: No Proc. Algorithm: Analyst Classic 4800 Bunching Factor: Noise Threshold: cps Area Threshold: cps 4600 Num. Smooths: 7 Sep. Width: Sep. Height: Exp. Peak Ratio: 5.00 Exp. Adj. Ratio: Exp. Val. Ratio: 3.00 RT Window: 30.0 sec 4200 Expected RT: 9.55 min Use Relative RT: No 4100 Int. Type: Base To Base Retention Time: 9.50 min Area: 4.70e+004 counts Height: 5.15e+003 cps Start Time: 9.30 min End Time: 9.79 min Time, min 9.50 Sample Name: "K5" Sample ID: "" File: "131121_Stettin_intestinal_samples.wiff" Peak Name: "UGT1A1*3(IS)" Mass(es): "510.1/753.8 amu" Comment: "" Annotation: "" Sample Index: 8 Sample Type: Standard Concentration: 1.00 ng/ml Calculated Conc: N/A Acq. Date: 11/21/ e4 Acq. Time: 3:55:50 PM 1.7e4 Modified: No Proc. Algorithm: Analyst Classic Bunching Factor: 1 1.6e4 Noise Threshold: cps Area Threshold: cps 1.6e4 Num. Smooths: 7 Sep. Width: 0.20 Sep. Height: e4 Exp. Peak Ratio: 5.00 Exp. Adj. Ratio: 4.00 Exp. Val. Ratio: e4 RT Window: 30.0 sec Expected RT: 9.56 min 1.4e4 Use Relative RT: No Int. Type: Base To Base Retention Time: 9.51 min Area: 1.48e+005 counts Height: 1.73e+004 cps Start Time: 9.33 min End Time: 9.79 min 1.4e4 1.3e4 1.3e4 1.2e4 1.2e4 1.1e4 1.1e4 1.0e Time, min 9.51 Method validation: specificity TYPVPFQR TYPVPFQR TYPVPFQR / / / / / / Intensity, cps Intensity, cps Intensity, cps Intensity, cps Intensity, cps Intensity, cps t=9.51 min t=9.52 min t=9.51 min t=9.50 min t=9.50 min t=9.51 min

39 Fallon et al. 2013, DMD Method validation: specificity

40 Method validation: linearity & range _Stettin samples_1-mixa.rdb (CYP2C19): "Linear" Regression ("1 / x" weighting): y = x (r = ) _Stettin samples_1-mixa.rdb (UGT2B15): "Linear" Regression ("1 / x" weighting): y = x (r = ) _Stettin samples_1-mixa.rdb (UGT1A1): "Linear" Regression ("1 / x" weighting): y = x (r = ) CYP2C UGT2B UGT1A A n a ly te A re a / IS A re a A n a ly te A re a / IS A re a A n a ly te A re a / IS A re a Analyte Conc. / IS Conc Analyte Conc. / IS Conc Analyte Conc. / IS Conc _Stettin samples_1-mixa.rdb (CYP3A4): "Linear" Regression ("1 / x" weighting): y = x (r = ) _Stettin samples_1-mixa.rdb (CYP2E1): "Linear" Regression ("1 / x" weighting): y = x (r = ) _Stettin samples_1-mixa.rdb (UGT2B7): "Linear" Regression ("1 / x" weighting): y = x (r = ) CYP3A CYP2E UGT2B A n a ly te A re a / IS A re a A n a ly te A re a / IS A re a A n a ly te A re a / IS A re a Analyte Conc. / IS Conc Analyte Conc. / IS Conc Analyte Conc. / IS Conc. Range for all proteins: nmol/l (LLOQ = 3.75 fmol per injection) Linearity proven for all 13 peptides (from at least N=6 calibration rows)

41 Method validation: accuracy & precision Protein Within-day data Between-day data Range Accuracy (%) Precision (%) Accuracy (%) Precision (%) (nmol/l) CYP1A CYP3A CYP3A CYP2B CYP2C CYP2C CYP2C CYP2D CYP2E UGT1A UGT1A UGT2B UGT2B Gröer et al. 2014, J Pharm Biomed Anal

42 Method validation: stability Parameter short-term (bench-top)stability (%) Post-preparative (rack) stability (%) Digestion stability (%) Protein 2 room temperature 24 5 C (autosampler) C CYP1A CYP3A CYP3A CYP2B CYP2C CYP2C CYP2C CYP2D CYP2E UGT1A UGT1A UGT2B UGT2B Gröer et al. 2014, J Pharm Biomed Anal

43 Digestion efficiency (%) Method validation: digestion 150 protein trypsin h 4 h 16 h 24 h CYP1A2 CYP3A4 CYP3A5 CYP2B6 CYP2C19 CYP2C8 CYP2C9 CYP2D6 CYP2E1 UGT1A1 UGT1A3 UGT2B7 UGT2B15 Gröer et al. 2014, J Pharm Biomed Anal

44 Accuracy of the entire analytical process Gröer et al. 2014, J Pharm Biomed Anal

45 Validation of LC-MS/MS assays accuracy and precision stability digestion efficiency species specificity Gröer et al. 2013, J Pharm Biomed Anal Gröer et al. 2014, J Pharm Biomed Anal

46 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

47 European inter-laboratory comparison Analysis of 10 identical human livers quantification of clinically relevant enzymes (N=13) and transporters (N=10) Use of different sample preparation procedures and MS techniques Mol Pharm. 2017; 4(9):

48 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

49 Critical aspects of mass spectrometry based proteomics - Application of the Method - Mol Pharm. 2017;14(9):

$fraction$

50 Critical aspects of mass spectrometry based proteomics - Application of the Method - Whole tissue lysates Cell fraction enrichment Mol Pharm. 2017;14(9):

Sample preparation: transporters disruption of frozen tissue tissue

Kit LC-MS/MS analysis (QTRAP5500) adjust concentration, denaturation,

(BCA) extraction of membrane proteins ProteoExtract Native Membrane Protein

transporters in plasma membrane determine function - Plasma membrane ~10% of

51 Sample preparation: transporters disruption of frozen tissue tissue homogenization cell lysis ProteoExtract Native Membrane Protein Extraction Kit LC-MS/MS analysis (QTRAP5500) adjust concentration, denaturation, reduction, alkylation, trypsin digestion determine protein concentration (BCA) extraction of membrane proteins ProteoExtract Native Membrane Protein Extraction Kit Limitation: - Isolation of crude membrane (CM) - Only transporters in plasma membrane determine function - Plasma membrane ~10% of CM risk of over- or underestimation transporter protein in crude membrane vs. transporter protein in plasma membrane

52 Critical aspects of mass spectrometry based proteomics - Application of the Method - 1 to 100-fold difference between groups Total protein vs. enriched cellular fraction comparing apples and oranges Wegler et al. 2017, Mol Pharm

53 Harwood et al. 2014, DMD

Sample preparation: enzymes tissue disruption and

$fraction centrifugation of supernatent 100,000 g cell$ $centrifugation (100,000 g) pellet: microsomal fraction$ an artificial product - Yield of microsomes is highly

an artificial product - Yield of microsomes is highly

) - Not surprising: highly variable data better: whole

54 Sample preparation: enzymes tissue disruption and homogenization centrifugation 9,000 g supernatent: S9 fraction centrifugation of supernatent 100,000 g cell debris supernatent: cytosolic fraction wash pellet centrifugation (100,000 g) pellet: microsomal fraction Limitation: - There are no microsomes in vivo, they are an artificial product - Yield of microsomes is highly variable (in dependence on prep.) - Not surprising: highly variable data better: whole tissue lysate, but resuspension of pellet, adjust concentration, denaturation, reduction, alkylation, trypsin digestion

$enriched cellular fraction comparing apples and oranges$

55 Critical aspects of mass spectrometry based proteomics - Application of the Method - 1 to 6-fold difference between groups Total protein vs. enriched cellular fraction comparing apples and oranges Wegler et al. 2017, Mol Pharm

$Critical aspects of mass spectrometry based proteomics - Application of the Method - Scaling enriched cell fraction total protein 3$

56 Critical aspects of mass spectrometry based proteomics - Application of the Method - Scaling enriched cell fraction total protein 3 to 30-fold difference between groups Scaling factor = [Protein] cell fraction Protein tissue lysate Enzymes (mean protein concentration)

$fractions whole tissue lysate$ is used - Enzymes and

57 Sample preparation for enzymes and transporters: FASP FASP: - Filter-aided sample preparation - Suitable for complex samples - No loss of subcellular fractions whole tissue lysate is used - Enzymes and transporters can be measured in the same sample - but: loss of sensitivity and enzymes function not assessable

58 Protein amount [fmol/mg tissue] Protein amount [fmol/mg tissue] Protein amount [fmol/mg tissue] Protein amount [fmol/mg tissue] Protein amount [fmol/mg tissue] Example: protein abundance of transporters in human intestine and liver ABCB ABCC ABCC D J1 J2 I C L 2 0 * * * * D J1 J2 I C L 5 0 D J1 J2 I C L ABCC ABCG * D J1 J2 I C L 0 * D J1 J2 I C L D: duodenum, J1: upper jejunum, J2: lower jejunum, I: ileum, C: colon, L: liver * below limit of quantification (2.7 fmol/mg tissue) Drozdzik et al., unpublished

59 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

proteomic analysis is normally determined by unspecific methods (e.g.

factors (ph, buffer, detergents, temperature et cetera) data observed by

60 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein content of cell or tissue lysate samples for proteomic analysis is normally determined by unspecific methods (e.g. Lowry, Bradford, BCA) Limitations: highly variable, influenced by many factors (ph, buffer, detergents, temperature et cetera) data observed by high-end MS are challenged by these simple methods Anal Chem Apr 21;87(8):

61 Digestion efficiency (%) The Achilles Heel of proteomics: protein digestion protein trypsin 2 h 4 h 16 h 24 h CYP1A2 CYP3A4 CYP3A5 CYP2B6 CYP2C19 CYP2C8 CYP2C9 CYP2D6 CYP2E1 UGT1A1 UGT1A3 UGT2B7 UGT2B15 Completeness of protein digestion is a prerequisite for assessing absolute protein abundance has to be determined for each protein and cannot be predicted QconCAT doe not overcome this issue as the resulting peptide chain is artificial SIL-proteins as internal standards may be an option (but: an in vitro model is needed, expensive) Best practice so far: characterize optimal digestion procedure (time / amount of trypsin), add marker protein as a surrogate for digestion (inter-sample variability can be minimized) Gröer et al. 2014, J Pharm Biomed Anal

62 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

63 ibaq values normalized to membrane protein ibaq values normalized to membrane protein Intestinal transporters: global proteomics total Membrane proteins ABC transporter SLC transporter Jejunum Ileum LS FHs Caco E E E-03 LS180 FHs 74 int Caco E E-04 Jejunum Ileum 2.00E E E E E E E E E E E E E+00

Global proteomics is a powerful tool for

used calculated absolute abundance is an

million Tweets per day People from Tibet

64 Absolute protein abundance by global proteomic data? Global proteomics is a powerful tool for high resolution identification of proteins (using many peptides, PTM also considered) no reference peptides are used calculated absolute abundance is an estimation (assumption: each peptide generates the same intensity) Limitations: low abundance proteins may be overseen, same limitations as targeted proteomics more than 500 million Tweets per day People from Tibet contribute daily about tweets 330 million monthly active users distribution around the world?

65 Critical aspects of mass spectrometry based proteomics - Application of the Method - Protein Biological / Clinical Question Protein sequence data (consider SNPs, PTMs) Sample preparation (fractionation, whole lysate) Protein determination / digestion Peptide Selection of appropriate peptides (in silico / wet-lab) Number of peptides / protein LC-MS/MS- Analysis Global or targeted proteomic analysis Selection of mass transitions (non- / scheduled MRM) Data analysis (peptides, MRMs, normalization) Validation of the method

66 Impact of different peptides, mass transitions and normalization Different peptides can result in different protein amounts 2-3 peptides should be used Different mass transitions of each peptides should be separately analyzed (no sum) Normalization: very different high variability of published data to unspecific protein (BCA) or specific protein to whole tissue / cell lysate or a certain fraction (membrane, microsomes) Tissue samples represent a mixture of different cell types housekeeper Fallon et al. 2013, J. Proteome Res.

$001 ABCC3 36% D J1 J2 I1 I2 C1 C2 C3 C4 ABCC2 25% no OATP1A2 Limitations Protein determination in crude membrane fraction$

67 PEPT1 50% small intestine Example: Intestinal transporter proteins ABCB1 8% OATP2B1 6% ABCC2 10% OCT1 8% OCT3 1% ABCC3 7% ABCG2 4% ASBT 6% PEPT1 1.5 OATP2B1 5% OCT3 2% 12% 1.2 OCT1 12% 0.9 ABCG2 3% colon protein expression (pmol/mg)1.8 ABCB1 (P-gp) ABCB1 5% p < ABCC3 36% D J1 J2 I1 I2 C1 C2 C3 C4 ABCC2 25% no OATP1A2 Limitations Protein determination in crude membrane fraction Only one peptide per protein No normalization on mucosal protein Drozdzik, Mol Pharm ;11:

the liver of 9 organ donors 80 70 60 50 40 30 20 10 0 ABCB1 D J1 J2 I C L Limitations considered Protein determination in whole

68 Protein amount [fmol/mg tissue] Example: protein abundance of transporters in human intestine and liver Aim: To analyze protein abundance of of clinically relevant metabolizing enzymes (N=13) and transporters (N=20) along the entire human intestine and in the liver of 9 organ donors ABCB1 D J1 J2 I C L Limitations considered Protein determination in whole tissue lysate (FASP) 2-3 peptides per protein Normalization to villin-1 (enterocyte marker protein) Drozdzik et al., unpublished

69 ketamine (pmol/mg min) Example: in vitro transport of ketamine OCT3 OCT1 OCT2 VC ketamine (µmol/l) Keiser, Mol Pharm. 2017

70 ketamine (pmol/mg min) ketamine (pmol/mg min) Example: in vitro transport of ketamine OCT3 OCT1 OCT2 VC ketamine (µmol/l) 6000 netuptake OCT3 OCT1 OCT2 Keiser, Mol Pharm ketamine (µmol/l)

71 ketamine (pmol/mg min) ketamine (pmol/mg min) Example: in vitro transport of ketamine OCT3 OCT1 OCT2 VC Total-Protein specific transporterprotein in total cell lysate ketamine (µmol/l) 6000 netuptake specific transporterprotein in crude membrane Specific transporterprotein in plasma membrane OCT3 OCT1 OCT2 Keiser, Mol Pharm ketamine (µmol/l)

72 ketamine (pmol/mg min) ketamine (fmol fmol OCTs -1 min -1 ) ketamine (pmol/mg min) proteinexpression in plasmamembrane (pmol/mg) Example: in vitro transport of ketamine OCT3 OCT1 OCT2 VC ketamine (µmol/l) 0 VC OCT 6000 netuptake OCT3 OCT OCT1 OCT OCT2 50 OCT ketamine (µmol/l) ketamine (µmol/l) Keiser, Mol Pharm. 2017

73 Summary & Conclusions LC-MS/MS-based targeted proteomics is a powerful tool to determine simultaneously absolute abundance of several ADME proteins High sensitivity, broad analytical range, high accuracy and precision and good reproducibility Limitation: completeness of protein digestion Sample preparation (e.g. fractionation) has a huge impact on the generated data (even higher than the used peptide) More than one peptide should be used (MRM analysis separately) Protein data should be generated from whole tisssue lysate (FASP) and not from enriched fractions Global proteomics is a complementary technique but provides per se no absolute data

74 Matthias Mann, 2008

75 LC-MS/MS-based proteomics: need for a consensus "Towards Reaching a Consensus on Using Quantitative LC-MS/MS Proteomics in Translational DMPK/PD Research" ISSX Workshop, 27th -28th September 2018, Cambridge, MA Wegler et al. 2017, Mol Pharm

Acknowledgements University Medicine Greifswald, C_DAT Clinical

University of Washington, Sweden Department of Pharmaceutics Per

Department of Pharmaceutics Prof. Dr. Jashvant Unadkat Dr.

Marek Drozdzik Marek Ostrowski Mateusz Kurzawski Pfizer Inc.

76 Acknowledgements University Medicine Greifswald, C_DAT Clinical Pharmacology Markus Keiser Diana Busch Janett Müller Anja Fritz University of Washington, Sweden Department of Pharmaceutics Per Artursson Christine Wegeler University of Washington, Seattle Department of Pharmaceutics Prof. Dr. Jashvant Unadkat Dr. Bhagwat Prasad University of Stettin, Poland Clinical Pharmacology Marek Drozdzik Marek Ostrowski Mateusz Kurzawski Pfizer Inc., Groton, CT Pharmacokinetics & Drug Metabolism Yurong Lai Larissa Balogh Charles Rotter

77 Thank you for your attention

Summary of Analytical Method for Quantitative Estimation of Fingolimod and Fingolimod Phosphate from Human Whole Blood Samples

Fingolimod Whole Blood Analysis Summary of Analytical Method for Quantitative Estimation of Fingolimod and Fingolimod Phosphate from Human Whole Blood Samples Study Detail: - Determination of Fingolimod